JPS5877010A - Musical signal communicator - Google Patents

Abstract

(57) [Summary] This bulletin contains application data before electronic filing, so abstract data is not recorded.

Description

DETAILED DESCRIPTION OF THE INVENTION The present invention relates to a technique for digitally encoding music signals for transmission and storage.

When digital transmission media and digital storage media are used for music signals, noise tolerance can be increased, so when using such media, it is easier to use than analog transmission technology and analog transmission technology that has been known for a long time. It's convenient. That is,
When using digital transmission technology and digital storage technology,
Noise inherent in the transmission or memory medium has a much smaller effect on the final reproduced music signal than in analog devices.

In typical known digital music signal encoding devices proposed to date, an analog input signal corresponding to a music signal is converted into a series of digital signals using an analog-to-digital converter. The digital warp pFi generally contains 14 to 16 bits representing the amplitude of the analog signal at a particular point in time. According to the well-known Nyquist sampler theory, the rate at which analog signals are sampled to form digital data words must be at least twice the highest frequency to be processed by the device. Therefore, a typical music signal encoder V in which the highest frequency to be reproduced is in the range of 15 Hz to 20 Hz.
In the device, the sampling rate must be at least 30 to 40 KHz.

Such known digital devices have the advantage that the data rates required to communicate musical signals and the data storage capacity required to store these musical signals in digital form are comparable to similar data communication rates required in analog form. It has a major drawback in that it is significantly larger than its storage capacity. For example, if you select 44 KHz as the sampling rate and 14 bits as the sample size, a stereo system using two channels will have 1
1.23 million bits per second are required. Under the same conditions, 1
If we use 6-bit samples, 140 per second
Ten thousand pits are required. These required data rates and corresponding storage requirements far exceed the capacities currently available in analog music devices.

Therefore, this type of digital music signal transmission and storage devices generally specifies the use of video storage technologies, such as video discs and video cassette recorders, in order to obtain the required data rates and storage capacities. Therefore, the cost and complexity of such digital devices far exceeds that of the analog devices they replace.

SUMMARY OF THE INVENTION It is an object of the present invention to provide a method for communicating and storing digital representations of musical signals in such a way that the data rates and storage capacity required are considerably lower. The above method is described in conjunction with an apparatus for receiving an analog music signal, converting it into digital data according to the principles of the present invention, and sending it to the digital booter or unit for converting it back to analog form. do. In accordance with another feature of the present invention, an analog music signal is received and converted to digital data in accordance with the principles of the present invention, which is stored in a digital data memory. This digital data memory is then connected to its associated unit, and the skeleton unit reads the digital data stored in the memory, converts it back to analog form, and outputs an analog music signal.

Another object of the invention is to provide a simple means of filtering the frequencies of a music signal by taking advantage of the digital representation of the music signal. According to one embodiment of the invention. For example, an analog music signal is converted into a series of digital data words P representing the amplitude of the analog music signal in a plurality of repeated sampling intervals. This one-point data word, organized in a manner called time domain, is converted to a frequency one-in by a known Fourier transformer. According to the series of the invention, discrete Fourier transform coefficients are computed from a digital coater word in the time domain for at least one set of frequencies. Each set of frequencies includes a dominant frequency and at least one frequency having an octave relationship thereto.

According to a further feature of the invention, the sets of frequencies are selected such that the main frequencies of each set are separated by the same percentage frequency spacing across the main octave. According to a further feature of the invention, each of these dominant frequencies corresponds to a different selected musical note. According to a further feature of the invention, a plurality of main frequencies are arranged within a predetermined frequency range for each note of the main octave.

In yet another embodiment of the invention, the complex discrete Fourier transform coefficients at selected frequencies are transformed from rectangular coordinate form having a real and imaginary part to polar coordinate form having a magnitude and angular part. Ru. Since the human ear is known to be less sensitive to the angle portion of information than to the magnitude portion of the information, the angle portion is represented by fewer bits than the magnitude portion. According to yet another feature of the invention, the magnitude part of the complex discrete Fourier transform coefficients is converted to logarithmic form in order to further reduce the data rate without significant loss of information.

In yet another embodiment of the invention, the effect of an iso-wavenumber filter is obtained by modifying the magnitude part of the complex discrete Fourier transform coefficients by an amount related to their corresponding frequency. If the magnitude portions are in proportional form, this modification is performed by multiplying each magnitude portion by a coefficient corresponding to the filter value of its associated frequency. If the magnitude sections are in logarithmic form, this modification is accomplished by adding 5 to each magnitude section a predetermined numerical value selected on the basis of its corresponding frequency.

These and other objects and features of the present invention will become apparent from the following detailed description taken in conjunction with the accompanying drawings.

Hereinafter, the present invention will be described in detail with reference to the accompanying drawings. 1st
The figure shows a preferred embodiment of the device according to the invention. FIG. 5 shows a preferred embodiment of a decoding device according to the invention. In the next communication channel, the equipment shown in Figure 1 is located at the transmitting station, and the equipment shown in Figure 5 is located at the receiving station. When storing music signals on a record P record or some type of magnetic recording medium such as a magnetic disk, magnetic tape or some portable memory device such as a cassette, the apparatus shown in FIG. 1 is used. , the music signal to be sent to the memory is encoded, and the digital data read from this portable memory is decoded using the apparatus shown in FIG.

FIG. 1 shows a preferred embodiment of an encoder device 100 according to the present invention. Analog input signal is analog input 101
The receiver is removed through the analog-to-digital converter 11
Sent to 0. This analog-digital converter 11
0 produces a series of digital data words corresponding to the analog input.

This series of digital data words from analog-to-digital converter 110 is sent to Fourier transform means 120. This Fourier transform means converts the 120 time r main digital data into frequency 9 main digital data at the selected frequency. This frequency domain digital data is sent to a rectangular coordinate to polar coordinate conversion means 130.

This orthogonal coordinate-polar coordinate conversion means I30 converts the complex discrete Fourier transform coefficients received from the Fourier transform means 120 from an orthogonal coordinate form having a real part and an imaginary part to a polar coordinate form having a magnitude part and an angle part. do. This complex number data in the form of polar coordinates is sent to the logarithmic conversion means 140. This logarithmic conversion means 140 converts the magnitude part of complex number data into logarithmic form. The complex number data in the form of polar coordinates converted in this way is sent from the logarithmic transformer=140 to the digital filter 150.・This digital filter 15
0 selectively modifies the magnitude part of each complex coefficient.

This fix is a selective frequency filter.

The output of the complex discrete Fourier transform coefficients from the digital filter 150 is stored in the memory means 160 or in communication with the frequency of the discrete Fourier transform coefficients of the particular complex number. The discrete Fourier transform means are stored for later retrieval and use. A communication channel 170 sends these complex discrete Fourier transform coefficients to another receiving and decoding device.

Analog-to-digital converter 110 operates to convert analog input data received at input 101 into a series of digital data words. This analog-to-digital converter 111 includes a low-pass filter 111, a sample-and-hold device 112, a clock 113, and a digitizer 114.
The cutoff frequency of is equal to the highest ripple frequency of the expected input signal sent to input 101. In a typical system where a music signal is sent to input 101, the cutoff frequency of low-pass filter 111 is 15 l-
4Z to 20Hz. Low pass filter 111
is used to prevent other high frequency signals from being sent to sample and hold device 112 that would interfere with the constructed analog signal. Sansol hold device 112 receives the filtered output from low pass filter 111. At time intervals determined by clock 113, sample and hold device 112 periodically samples the amplitude 't of the analog signal sent to it and holds this sample until the next sample is taken and issued. do. The sample hold 0 device 112 typically takes the form of an r-type amplifier that drives an output compacitor, the output compacitor having a relatively large time constant compared to the sampling interval. connected to the circuit. The sampled magnitude signal from sample-and-hold device 112 is sent to digitizer 114 along with the signal from clock 113.
14 forms a digital data word having a value corresponding to the magnitude of the sampled signal.

2, 5, and 4 illustrate various embodiments of the digitizer 114 shown in FIG.

FIG. 2 shows a digitizer with a ramped voltage panel.

The sample voltage from sample and hold device 112 is sent to the non-inverting input of comparator 103. The output of ramp voltage generator 104 is sent to the inverting (-) input of comparator 103. Signals from clock 113 start and reset ramp voltage generator 104 and set fritz flop 105. When the fritz flop 105 is set, the Noah gate 10
A logic high signal is applied to one input of 7. Therefore, the NOR gate 107 converts the signal from the oscillator 106 into the counter 10.
It works as an inverter to pass to 8. The frequency of oscillator 106 is also chosen to be much higher than the frequency of clock 113. The voltage from ramp voltage generator 104 starts from a small value and increases linearly, but when this voltage crosses the voltage from sample and hold device 112, comparator 1
The output of 03 resets flip-flop 105. Therefore, Noah f-) 107 can no longer pass the output of oscillator 106 to counter 10g. At this time, counter 1
The counts accumulated in ramp voltage generator 104
It is directly related to the time it takes for the voltage from to reach the sampled voltage, which is in turn proportional to the sampled voltage. The counter 108 is started or reset by a signal from the clock 113 so as to be synchronized with the sampling timing of the sample/hold device 112.

FIG. 3 shows a tracking counter type digital type. The voltage from the sample/hold device 112 is measured by the comparator 103.
is sent to the non-inverting (g) input of The output of comparator 103 is sent to the up/down control input of up/down counter 115. The °up/down counter 115 is driven by the oscillator 106. The output of up/down counter 115, which is the desired digital data output, is sent to digital-to-analog converter 116. The digital-to-analog converter 116 is
generates an analog voltage that is proportional to the voltage
This analog signal is applied to the inverting input of the comparator 103.
Equipped with a feedback loop that makes the count accumulated in the up/down counter 115 proportional to the sampled voltage ◎Up/down counter 1
If the count of 15 is too small, the denotator inverter 116 also produces an output signal that is smaller than the sampled voltage. In this case, the comparator 103
The down counter 115 is placed in count up mode.

On the other hand, if the count of up/down counter 115 is too large, digital-to-analog converter 116
The output of the voltage is greater than the sampled voltage. In this case, the output of the comparator 103 puts the up/down counter 15 into countdown mode. In this way, the feedback loop causes the up/down counter 115 to reach the desired digital 1 data power. Make sure to count in the opposite direction. Up/down counter 115 samples
Clock 113 to be synchronized with hold device 112
The operation is reset or started by a signal from the Since the i location shown in FIG. 3 typically toggles the least significant bit between 1 and 0, up/down counter 115 contains more than the bits needed for the desired digital data word 0. Note that it is common for one more bit to be included.

FIG. 4 shows a continuous approximation type digital type.

The sample voltage from sample-and-hold device 112 is applied to the non-inverting input of comparator 103. Similar to the example shown in FIG.
6 supplies an analog voltage to the comparator 1030 inverting ←) input. The output of comparator 103 is sent to the up/down input of sequencer 117. Sequencer 117 presets the count of presettable counter l18 starting with the most significant bit and then ending with the least significant bit. Initially, the count stored in this presettable counter 118 is completely zero, so the output from the digital-to-analog converter is less than the sample voltage. The most significant bit is then set to 1, so that the digital-to-analog converter 116 generates a mid-range voltage controlled by the digital data word stored in the presettable counter 118. If the voltage output from digital-to-analog converter 116 is still less than the sample voltage, then the output from comparator 103 is unchanged and the desired value of the most significant bit is one.

Digital to analog converter! When the output from comparator 103 becomes greater than the sample voltage, the output from comparator 103 is inverted and therefore the most significant bit becomes zero. The second most significant bit is then set to 1, and comparator 10
A comparison is made again by 3 to determine whether the correct value for this bit is zero or one. This process continues for all pits of resettable counter 118 until each of these bits is thus set. Sequencer 117 clocks clock 1 so as to be synchronized with sample and hold device fll 12.
The operation is reset or started by a signal from 13.

The digital data word is sent to Fourier transform means 120. The Fourier transform means 120 converts the time domain data received from the analog-to-digital converter 110 into frequency domain data at a particular set of selected frequencies. The Fourier transform means 120
21, 122 and (OFT). Here, this i is a set of discrete A standard Fourier transformer is required. These discrete Fourier transformers perform a fast Fourier transform on sampled data in the time domain. The Fourier transform is a mathematical technique that utilizes matrix operations to simplify the calculations required to calculate the Fourier transform coefficients on sampled data in the time domain. This mathematical technique is dedicated to the amount of data calculation... - Dounia? Examples also include! The invention is equally applicable to embodiments using program-based digital computers. In the application discussed here, a set of digital data words corresponding to approximately 50 milliseconds of analog input are applied to the input of a discrete Fourier transformer. In a known typical usage, the output of a discrete Fourier transformer is a set of complex discrete Fourier transform coefficients corresponding to a set of proportionally spaced frequencies, where the spacing is Corresponds to the reciprocal of the length of data represented by a data block. Therefore, for example,
As longer data word blocks are sampled, greater frequency analysis power can be obtained, but the amount of computation required to obtain data with this greater frequency analysis power also increases accordingly. In the use of the discrete Fourier transformer discussed here, rather than complex discrete Fourier transform coefficients being computed for each of these proportionally spaced frequencies, Discrete Fourier transform coefficients are calculated only for the dominant frequency, which is the lowest frequency for which the digital Fourier transform can be solved. This is related to the length of the data sample and also to the frequency which has an octave relationship to the dominant frequency. Therefore,
Discrete Fourier transform coefficients are calculated only for the dominant frequency F and frequencies matching Equation 2F (where n is an integer). Computing discrete Fourier transform coefficients for only these frequencies slightly modifies the fast Fourier transform calculation method to eliminate the matrix operations required to compute discrete Fourier transform coefficients corresponding to other frequencies. It is done by

The Fourier transform means [20 is shown to include a plurality of discrete Fourier transformers 121, 122 and 123 in order to calculate discrete Fourier transform coefficients for multiple sets of frequencies with an octave relationship. Thus, each discrete Fourier transformer operates on a different number of consecutive time domain samples from the analog-to-digital converter 110 and has a set of dominant frequencies within the dominant octave. Create discrete Fourier transform coefficients for the frequency of .

The operation of the discrete Fourier transform means is generally illustrated in FIG. Curve 301 corresponds to the music signal sent to input 101, which is in the time domain. An analog-to-digital controller forms a digital data word corresponding to the amplitude of curve 301 at each time indicated by 110 samples 302. These digital data words are sent as samples designated group 303 to the discrete Fourier transform means. As mentioned above, the length of the sample group determines the frequency resolution power of the discrete Fourier transform. Curve 304 represents the Fourier transform of the analog waveform shown by curve 301;
- It is understood that the same information represented in the time domain in 01 is represented in the frequency domain. sample 30
When the sample group 303 of 2 is sent to a discrete Fourier transformer based on a known technique, a plurality of discrete Fourier transform coefficients 30 whose frequencies are separated by the reciprocal of the time length of the sample group 303 are generated.
5 is formed. Note that each of these discrete Fourier transform coefficients is a complex number that corresponds to a frequency domain sample at a particular frequency of the analog wave shown by curve 301.

According to the invention, only a sub-set of these discrete Fourier transform coefficients is calculated. This subset consists of selected frequencies 30
6. According to Fuhakaido, this selected frequency 306 includes a number of dominant frequencies and a group of frequencies with an octave relationship to each of them.

The complex discrete two-dimensional transform coefficients at the selected frequency are sent from the Fourier transform means 120 to the orthogonal dust coordinate to polar coordinate transform means 130. Fourier transform means 120
Note that the complex numbers formed by are in Cartesian form. That is, each of these complex numbers is represented by a real part and an imaginary part orthogonal to the real part. The orthogonal coordinate-polar coordinate conversion means 130 converts each coefficient of these complex numbers from an orthogonal coordinate representation to a polar coordinate representation. In polar coordinate representation, each of these complex numbers has a magnitude part and an angle part. A complex number in rectangular coordinate form is converted to a complex number in polar coordinate form using the following equation.

where M is the magnitude part of the polar coordinate representation, A is its angular part, and R is the real part of the rectangular coordinate representation, and l
is its imaginary part.

The orthogonal coordinate-polar coordinate conversion means 130 includes a multiplier 1
31 and 132, adder 133, and address means 13
4, a square root lookup table 135, a divider 136, an address means 137, and an arctangent lookup table 138. When each of the complex discrete Fourier transform coefficients is received by the Cartesian to polar coordinate transformation means 130, its real part R is sent to the dual power of the multiplier 131, and R
2 is calculated. Similarly, the imaginary part 1di is sent to both sides of the ripple layer 132, and r and -1rE are calculated. The outputs of multipliers 131 and 132 are sent to both sides of adder 133. The result of this addition is sent to the address means 13, 4, and the data to be read from the square root lookup table 135 is determined. In particular, address means 134 forms an address based on signals received from adder 133, and data values are accessed from square root look-up table 135 to read the square root of the number sent to address means 134. Addressing means 134 determines that the output from adder 133 is one of the square root values stored in square root lookup table 135.
If there is not an exact match, a rounding function may be included in the form of accessing the nearest square root q@o stored in the square root lookup table 135. The output of the square root value from the square root lookup table 135 corresponds to the magnitude part of the complex number.

The complex angular part in polar coordinate form is calculated using a divider 136, an addressing means 137 and an arctangent look-up table 138.

The imaginary part 1 and the real part R of the Cartesian coordinates are sent to a divider 136 to form T/R. This is address means 13
7, this address means is sent to address means 134
Arctangent lookup table 1 as well as interaction with square root lookup table 135
Controls reading of data from 38. That is, the address means 137 selects from the arctangent lookup table 138 an address in which data corresponding to the arctangent of the successive value output from the divider 136 is stored. Therefore, the output of the arctangent lookup table 138 is the complex angle part in polar coordinate form. As described below, the angular part of a complex number expressed in polar coordinate form represents the magnitude part of the complex number expressed in polar coordinate form or the real or imaginary part of a complex number expressed in rectangular coordinate form. The number of bits used for is also expressed by a small number of bits. For this reason, the angles are quantified in a relatively rough manner and therefore there is no need to perform precise calculations to select the appropriate angles. First of all
In addition, the sign pits of the real and imaginary parts correspond directly to the quadrant in which the angular part exists, h', and thus to the top two pits of the angular part. Furthermore, the division performed by divider 136 is performed only on the top few pits of imaginary part 1 and real part R, without affecting the selected angular part. Alternatively, the absolute value of the magnitude of the most significant bit of the imaginary part 1 and the real part R may be compared to derive the least significant bit of the angular part. These or other similar methods can be used to reduce the amount of computation required to calculate the angle.

The magnitude and angle parts of each complex Fourier transform coefficient from the Cartesian to polar transform means 130 are sent to the input of the logarithmic transform means 140 . The logarithmic transformation means 140 operates only when it receives the magnitude part of a complex Fourier transform coefficient. This thick part M is sent to addressing means 141 which addresses a logarithmic look-up table 142. Addressing means 141 and logarithmic lookup table 142 operate similarly to addressing means 134 and square root lookup table 135 described above.

The address in logarithmic lookup table 142 where the data corresponding to the logarithmic value of the digital data word was stored is selected. Therefore, the output of the logarithm look-up table 142, and thus the output of the logarithm conversion means 140, is 1. It corresponds to the logarithm of the magnitude part of the complex Fourier transform coefficients sent to the means.

Conversion from rectangular coordinate format to polar coordinate format is effective in terms of the human ear's responsiveness to sound. The human ear is relatively insensitive to phase angle.

Therefore, the angular part of a complex discrete Fourier transform coefficient represented in polar coordinate form can be represented with a relatively small number of pits without loss of fidelity. In addition to this,
The human ear responds logarithmically to the amplitude of sound. Therefore, even if the same range of sound amplitude is encoded by representing the logarithm of the magnitude part with fewer pits than the required bits, the fidelity will be low. You won't lose it. A signal that requires a 14-bit data word to faithfully represent in the time domain can use only a 10-bit data word in the frequency domain, 6 bits for magnitude and 4 bits for angle. It turns out that this same signal can be represented with the same fidelity.

The output of the logarithmic conversion means 140 is sent to a digital filter 150. Operates as a digital wave number filter. digital filter 150
comprises an adder 151, a counter 152, address means 153, and a frequency coefficient lookup table 154. Note that it is only the logarithm+represented magnitude portion of the complex Fourier transform coefficients that is acted upon by the r-phthal filter 150. This logarithmic magnitude portion is sent to one input of adder 151. The other input of adder 151 is derived as follows. Logarithmic conversion means '140 and digital filter 1
counter 15 synchronized to the data transfer rate between
2 drives the address means 153. The output of the Fourier transform means 120 is organized such that the complex discrete Fourier transform coefficients output therefrom are in a predetermined repeating order based on their corresponding frequencies. counter 152
The counting capacity of the counter 1520, which is synchronized to the data rate of this device, is therefore equal to the number of frequencies selected for calculating the complex discrete Fourier transform coefficients. The address means 153 receives the count of the counter 52, and the current receiver represents the frequency corresponding to the complex Fourier transform coefficient taken. Select the address in the lookup table 154 for the coefficient.The output of this lookup table 154 is sent to the other input of the adder 151, so that the adder 151 receives the magnitude of the complex 7+7E transform coefficient. The output of the adder 151 is the output of the digital filter 150. Adding the numerical value forming the magnitude part of the logarithm means multiplying by a predetermined coefficient forming the magnitude part expressed in proportional form. Please note that it is the same as

The output of digital filter 150 is sent to either memory 160 or communication channel 170.

When device 100 is used to store digital data for later playback, the data output from digital filter 150 is sent to memory 160. memory 16
0 may be a permanent memory such as a permanently fixed disk with a modified groove, or may be a writable memory such as a magnetic recording medium. The magnetic recording medium may be a computer magnetic disk, a reel-to-reel tape, a cartridge tape, or a cassette tape. When this encoding device 100 is used for communication, the digital encoder 150. FIG. 5 shows a decoding apparatus for decoding digital music signals in accordance with the present invention. This decoding device 200 includes a digital filter 2
10, an antilog conversion means 220, a polar coordinate-orthogonal coordinate conversion means 23G, an inverse Fourier transformation means 240, and a digital-to-analog conversion means 250. Digital filter 210 receives digital data input at input 201 from a suitable reading means of a memory, such as memory 160 shown in FIG. 1, or from a communication channel, such as communication channels 1-70 shown in FIG. I will take it. Digital filter 210 operates as a frequency filter in much the same way as digital filter 150 shown in FIG. The filtered digital data word from digital filter 210 is sent to antilog conversion means 220 .

This antilog conversion means 220 operates in the opposite manner to the operation of the logarithm conversion means 14'0 shown in FIG. 1 to convert the logarithmic magnitude data into proportional form. Convert to data. The digital data converted in this way is sent to the polar coordinate-orthogonal coordinate conversion means 230. This polar coordinate-orthogonal coordinate conversion means 230 converts the complex Fourier transform coefficients in the form of polar coordinates received from the antilog number conversion means 220 into the form of rectangular coordinates. This operation is the reverse of the operation performed by the Cartesian to Polar conversion means 130 shown in FIG. The complex Fourier transform coefficients in the form of rectangular coordinates from the polar coordinate to rectangular coordinate transform means 230 are sent to the inverse Fourier transform means 240 . This inverse Fourier transform means 2
40 transforms the discrete Fourier transform coefficients at the selected frequencies received from the polar to rectangular transform means 2 and 30 into the time domain. Inverse Fourier transform means 240
The output of is a set of time domain digital data words, each digital data word having a value corresponding to the magnitude of the analog signal. Inverse Fourier transform means 240
performs an operation opposite to that of the Fourier transform means 12G shown in FIG. The time domain digital data word from the inverse Fourier transform means 240 is sent to the digital-to-analog converter means 250. The digital-to-analog conversion means 250 receives time domain digital data word samples from the inverse Fourier transform means 24Q and generates a corresponding analog signal. This operation is the reverse of the operation of analog-to-digital converter 110 shown in FIG. Digital-to-analog conversion means 250 produces an analog output at output 202.

Digital filter 210 operates similarly to digital filter 150 of FIG. 1 described above. Digital filter 2
10 receives a digital data word at digital data input 201. This digital data word is stored in memory 16
These digital data word inputs, which may be from a suitable readout device of the memory, such as 0, or may be received from a communication channel, such as communication channel 170, are input to one input of adder 211. sent to. Counter 212 is synchronized with the rate of reception of data words at digital data input 201. ``This synchronization is achieved by K by synchronizing with the reading means of memory 160, or by receiving and detecting a synchronization signal associated with the digital data word on communication channel #170. Ru. The count of counter 212 is sent to address means 213. As described above for digital filter 150, address means 213 selects the address in frequency search table 214 that corresponds to the currently received frequency. This frequency coefficient is sent to the other input of adder 211. In the same way as described above regarding the digital filter 150, the addition by this adder 211 results in %
7'' The magnitude part of the complex Fourier transform coefficient of the polar coordinate form n sent to the digital data manual 201 is corrected. The output of the adder 211 is the output of the digital filter 210.

The antilog conversion means 220 takes the corrected complex Fourier transform coefficients in the form of polar coordinates formed by the digital filter 210 and converts the magnitude part expressed in logarithms back into the proportional form. The conversion means 220 comprises address means 221 and an antilog look-up table 222. As already mentioned for the address means and look-up table combination of WrJ, this address means 221 forms an address corresponding to the magnitude part of the complex number taken. Antilog look-up table 222 stores in this location the data word corresponding to the antilog of the magnitude section taken. This antilog digital data word is read from the antilog lookup table 222 when an address is sent to the antilog lookup table 222 by the addressing means 221. The output of the antilog lookup table 222 is the output of the antilog conversion means 220.

The polar coordinate-orthogonal coordinate conversion means 230 is the antilog number conversion means 22
2 and converts complex numbers in polar coordinate form to rectangular coordinate form. The polar coordinate-orthogonal coordinate conversion means 230 includes an address means 231, a cosine lookup table 232, and a multiplier 23.
3, address means 234, and a sign look-up table 235.

The angle part of the complex number expressed in polar coordinate form is sent to address means 231. The address means 2° 31 selects the appropriate address corresponding to the digital data word of the received angle and applies it to the cosine look-up table 232. 231 outputs a digital data word corresponding to the cosine of the angle taken. This cosine ring is sent to one of the multipliers 233. The magnitude part M of the complex number is sent to the other input of this multiplier 233. Therefore, the output of the multiplier 233 is determined by formula 13 R+=Mcos^,
It is the real part of a complex number in Cartesian coordinate form.

Similarly, the complex angular part ^ in polar coordinate form is also sent to addressing means 234 so that a digital data word corresponding to the sine to this angular part is output from the sign look-up table 235 . This sine ring is sent to one input of a multiplier 236, and the magnitude part M is sent to the other input of this multiplier. Therefore,! The output of the lute derailleur 236 is expressed by the formula 1 = M sin
A K, the imaginary part becomes 1.

FIG. 6 shows a polar coordinate-orthogonal coordinate conversion/antilog number conversion means 26 that can be used in place of both the antilog number conversion means 220 and the polar coordinate-orthogonal coordinate conversion means 2300 shown in FIG.
It shows 0. This polar coordinate-orthogonal coordinate conversion and antilog number conversion means 260 includes an addresser 261, a lookup table 262, an adder 263, an addresser 264,
A lookup table 265, an addresser 266,
It includes a lookup table 267, an adder 268, an addresser 269, and a lookup table 270. The complex angular part expressed in polar coordinate form is sent to both addressers 161 and 166. Addresser 1
61 selects the address in lookup table 262 that corresponds to the angle taken by this receiver. The data stored at this location corresponds to the logarithm of the cosine of the angle ^. Similarly, addresser 266 selects the corresponding address in lookup table 267 for the received angle. The data stored at this location corresponds to the logarithm of the sine of the angle. Data from lookup table 262 is passed to one input of adder 263, which also receives the logarithm of the magnitude part M. Similarly, the output of lookup table 267 is sent to one input of adder 268, which also receives the logarithm of the magnitude part M.

The sum output from 7/-263 is sent to addresser 264, and addresser 264 of and- is sent to lookup table 2.
Select an appropriate address within 65.

The data at this selected address in lookup table 265 corresponds to the antilog of the output from adder 263. Similarly, the sum output from adder 268 is sent to addresser 269 and lookup table 27
An appropriate position within 0 is selected. As in the case of look-up table 26.5, look-up table 2'10 records data corresponding to the antilog of the sum output from Ame-268, thus polar-cartesian. The coordinate transformation and antilog number transformation means 260 uses the formula R=z number (log
M + log cosA)) (Thus, the real part R is calculated.Similarly, the polar coordinate-orthogonal coordinate conversion and antilog number conversion means 2
60 is the formula l; antilog (log M+ log , lnA
) to calculate the imaginary unit. When calculating orthogonal coordinates using K in this way, the polar coordinate-orthogonal coordinate conversion means 2
30 multipliers 233 and 236, the polar coordinate-orthogonal coordinate conversion and antilog number conversion means 260 includes an adder 26.
3 and 268 will be used. Such replacement greatly reduces the amount of calculation required to calculate orthogonal coordinates. Lookup tables 232 and 235 for trigonometric functions and lookup table 2 for logarithms of trigonometric functions
Since the only difference between 62 and 267 is the specific data in these tables, this difference does not require any additional calculations.

The inverse Fourier transformer 2) 40 receives the complex numbers in the form of rectangular coordinates from the polar to rectangular coordinate transformation and antilogarithmic transformation means 230 and transforms this data into the time domain. Here, since calculation is required to convert frequency domain input data to time domain, the Fourier transform means 120
As in the case of IC1, this inverse Fourier transform means 240 is
A number of discrete Fourier transformers represented by discrete Fourier transformers 241, 242 and 243 are provided. An important and useful feature in the mathematical analysis of discrete fast Fourier transforms is that this mathematical technique is completely reversible. That is, when converting data sampled in the time domain to the frequency domain,
The same matrix operations can be used both to convert data sampled in the frequency domain to the time domain. Therefore, each discrete Fourier transformer 241, 242 and 243 corresponds to the discrete Fourier transformer 121, 122 and 12 shown in FIG.
3 and performing the same mathematical function, the output of this group of discrete two-to-thousand converters consists of a set of digital data, each corresponding to the magnitude of the analog signal at a selected time. Word. This set of dental data words corresponds to the digital data word output from analog-to-digital conversion means 110, except for the frequency filtering effects of digital filter 150 and digital filter-210.

The time domain digital data word from the inverse Fourier transform means 240 is sent to the digital-to-analog converter means 250. The digital-to-analog conversion means 250 converts the received digital data words into analog signals and outputs them on the analog signal output 202. The digital-to-analog conversion means 250 includes a register 251;
A number of current generating means represented by current drives 252, 253 and 254, an analog summing amplifier 255, and a Ronos filter 2s6. ! - is in stock. Each digital data word received is sent in parallel to the receiver/meta 251, where the entire digital data word is stored. Each individual bit of the digital data word is sent to one of a group of current generating means indicated as current drives 252, 253 and 254. The state of the corresponding bit stored in register 251 determines whether each current drive is activated or deactivated.

Note that the amount of current produced by each current drive when it is active corresponds to the particular digit being represented. Therefore, current drive 2
52 generates 1 unit of current, current drive 253 generates 2 units of current, and current drive 254 generates 2N units of current. where N is the number of pits in the digital data word stored in register 251. Each of the currents generated by current drives 252, 253 and 254 is sent to an analog summing amplifier 255. The magnitude of the output of this analog summing amplifier 255 is
Corresponds to the value of the digital data word stored in register 251. This analog output is low-pass filtered 25
6, this filter serves to filter out switching transients between long-lived digital data words sent to register 251; the output of filter 256 is output 2.
Sent to 02. This output is digital filter 15G
and excluding the filtering action of the digital filter 210,
substantially corresponds to the analog input sent to input 101.

FIG. 8 shows selected frequencies of the Fourier transform coefficients in the frequency domain for the Fourier transform means 120 shown in FIG. 1 and the inverse Fourier transform means 240Iζ shown in FIG. It is known that the human ear generally distinguishes musical tones based on the difference in frequency between the center and the center.

Therefore, the human ear has almost the same analytical power for sound in each octave. By making good use of this limited analytical power in the human auditory system, we can calculate the number of frequencies that require discrete Fourier transform coefficients. One general idea of the present invention, which can significantly reduce the number of times, is to form discrete Fourier transform coefficients in sets of octave-related frequencies.

FIG. 8 shows the frequency dactre on a logarithmic scale, ie each octave occupies the same proportional space as each of the other octaves. This logarithmic scale approximates the analytical power of the human ear described above. This frequency spectrum comprises 2, 404 within the main octave. Each frequency pair consists of a main frequency in the main octave and a sub-frequency in each sub-octave,! -1”
tA,, -c i2s, :(j!5aa411d!:t
? #'-, KA in 401, be careful of 6 members 1i&i,j
The ob 1 frequency 411 has a side frequency □ wave number 412 in the sub-octave 402, and a side frequency 41 in the sub-octave 403.
3, and side frequencies such as side frequency 414 in minor octave 404 are combined. Main frequency 421 also falls within main octave 401 . This main station wave number 421 has a side frequency 422.423 which has an octave relationship with it.
and 424. Similarly, the main frequency 431 includes frequencies 432, 433 and 4 which have an octave relationship.
34 are combined. Main frequency 4゛41 and side frequency 44
A similar relationship holds between 2% 443 and 444. The main frequencies 411, 421, 431 and 441 are equally spaced from each other by /4 centimeters within the main octave 401. Therefore, the side frequencies within each sub-octave also span that respective sub-octave? are separated by the same frequency interval. Therefore, these selected frequencies for calculating the discrete Fourier transform coefficients are spaced equal to and equal to /mother-cent frequency over the entire frequency spectrum. 'The frequency dispersion at this same centezoon over the entire frequency spectrum is approximately equal to the response of the human auditory system to sound.'According to the idea of the present invention, the higher the octave, the more the frequency dispersion is calculated for that frequency. The number of discrete Fourier transform coefficients is smaller than previously known, but since the human auditory system is relatively insensitive to frequencies in higher octaves, some frequencies may be omitted. However, the fidelity of the reproduced musical tones is not reduced.

FIG. 9 shows a preferred embodiment of selecting the dominant frequency for the middle dominant octave. It should be understood that in each minor octave (not shown in FIG. 9), the side frequencies associated with the main frequencies shown in FIG. 9 are similarly distributed.
The figure shows a raw octave 401 and a group of main frequencies distributed across this main octave. In the embodiment shown in FIG. 9, the twelve main frequencies 501-512 are equally distributed over the main octave. In this case, 12 main frequencies 501 to 512
Each corresponds to the frequency of 12 musical tones in the main second tarp 401. Therefore, the main frequency 501 corresponds to the musical note ^,
The main frequency 502 corresponds to a semitone lower than the note B, the main frequency 503 corresponds to the note date, and so on, and so on, the main frequency corresponds to each note over the entire main octave. Since each of the main frequencies 501-512 has a number of side frequencies in various octaves associated with it, we compute discrete Fourier transform coefficients for the frequencies corresponding to each note over the entire spectrum. Note that the selection of frequencies is shown in FIG. 2 As 1 the human auditory system only perceives a percent jl wavenumber difference, and 2nd, these selected frequencies to calculate the discrete Fourier transform coefficients are the largest amount in encoding the music signal. This choice of frequency is particularly effective because the energy of exactly corresponds to the expected frequency.

10 and 11 show a portion of the main octave corresponding to notes B and CK, illustrating an improvement in the selection of frequencies for calculating the discrete Fourier transform coefficients. Instruments do not form a clear frequency envelope that exactly matches the frequency of the note being played, but rather a somewhat broad spectrum that generally corresponds to frequencies around the frequency corresponding to the note being played. It is generally known to form. In particular, pianos and pipe organs use two or more frequency generators for each note, with one or more frequency generating elements tuned lower and higher than the frequency corresponding to the note. By having one or more frequency generating elements carefully tuned to create a wide frequency spectrum. For such instruments, the frequency selection shown in FIG. 8 does not transmit all the useful information needed to reproduce the musical signal with high fidelity. The two choices shown in FIGS. 10 and 11 allow such wide music signals to be sampled and reproduced more accurately.

FIG. 10 shows a portion of the main octave containing notes B and C. The potential energy around the note 80 frequency is shown by an envelope 601.

The potential energy around the frequency of note C is envelope 6
Note that for each of the illustrated notes B and C, there are five selected frequencies, denoted by 02. The frequencies corresponding to note B are the main frequency 610 which is equal to the frequency of this note, the main frequency 611 which is lower than the frequency of this note, and the main frequency 61 which is higher than the frequency of this note.
, 2.

Similarly, there are also three main frequencies corresponding to note C.

These are a frequency 620 equal to the frequency of this note, a main frequency 611 lower than the frequency of this note, and a main frequency 622 higher than the frequency of this note. Assume that a group of three main frequencies are distributed within the potential energy envelope of each note in the main octave, as shown for the two main notes 8 and C in FIG. The main octave therefore contains 56 main frequencies.Each of these main frequencies is associated with a group of side frequencies, so that each note has discrete Fourier transform coefficients over its entire frequency spectrum. Calculate the frequency by 3
You will have one set of each.

Figure 11 is a diagram similar to Figure 10, but with notes 6 and C.
A pair of dominant frequencies is shown for each of the following. FIG. 11 shows a potential energy envelope 701 corresponding to note 8 in the main octave and a potential energy envelope 702 corresponding to note C in the main octave. Two main frequencies are combined for both notes 8 and C.

Note B has a combination of a main frequency 711 lower than that frequency and a main frequency 712 higher than that frequency. Similarly, note C has a main frequency of 7 lower than that frequency.
21 and a main frequency 722 higher than that frequency. Each of the 12 notes 9 in the main octave is associated with two main frequencies, and the main frequencies are 24
become individual. Similar to the frequency selection shown in Figure 8,
Sub-frequencies that are illustrated and related to the main frequency of interest exist throughout the frequency spectrum. According to the frequency selection method shown in FIG. 11, each note in the spectrum is therefore associated with two frequencies for which the discrete Fourier transform coefficients are calculated.

The Fourier transform means 120 can be constructed based on the aforementioned principle of selecting the frequencies at which the discrete 7-lier transform coefficients are calculated so as to minimize the amount of data computation required. This is the discrete Fourier transformer shown in FIG.

This can be achieved by providing discrete Fourier transformers such as 122 and 123, one for each main frequency. Each of these discrete converters operates on a different number of digital data words from analog-to-digital converter 110. As mentioned above, the lowest analyzable frequency of the output from the discrete Fourier transformer is related to the length of time sampled and thus the number of samples sent to the discrete Fourier transformer. In general, a discrete Fourier transformer is capable of producing discrete Fourier transform coefficients at the lowest frequency that can be analyzed and at all frequencies that are integral multiples of this frequency. A discrete Fourier transformer is provided for each dominant frequency, with each discrete Fourier transformer being responsive to a predetermined number of digital data word temples from analog-to-digital converter 110 corresponding to that dominant frequency. This eliminates the need for a single discrete Fourier transformer to operate on a large number of data word samples to provide frequency analysis power corresponding to each frequency within the main octave. Furthermore, each of these discrete Fourier transformers can also form discrete Fourier transform coefficients at sub-frequencies that have an octave relationship to their respective main frequency. The calculation of the discrete Fourier transform coefficients at frequencies other than those that have an octave relationship to the principal frequency of a particular discrete Fourier transformer is
The technique described above, which is prevented by excluding the matrix operations necessary for calculating these coefficients, will become clear from the following description of a specific example. Figure 9 shows that with >l, q, the recent dominant frequencies needed to calculate the discrete Fourier transform coefficients are 27,5
Note that it is in Hz. Assuming a sample data rate of 44 KHz, it is necessary to use 1600 samples to analyze this frequency, so a basic data clock consisting of 1600 digital data words is 7-rie conversion means 12
Sent to 0.

Discrete Fourier transformers for other higher dominant frequencies are responsive to a series of fewer consecutive digital data words than the 1600 data words that make up this basic data block. For example, to analyze the frequency of 3 0 , 9 Hz corresponding to note B, 1424 consecutive digital data words r are used. Thus, a discrete Fourier transformer corresponding to a dominant frequency of 30,9 Hz responds to 1424 consecutive digital data waves P from the analog-to-digital converter 110 during each cycle of 1600 digital data waves R. and must not respond to any other data words. This set of 14'24 consecutive digital data words lies somewhere within 1600 data word locks. Similarly, 3 corresponding to note C
A discrete Fourier transformer with a main frequency of 2.7 Hz must respond to 1346 consecutive digital data words in a data block of 1600 data words. If we use the frequency selection method shown in Figure 10 and assume that the deviation from the center frequency is about 0.5 inches, the frequency 611 is about 3 Q, 7 Hz.
, which requires 1431 samples. The frequency 610 is approximately 50-a 9 H2, which is 1424
97 pulls 7 are required, and the frequency 612 is 31, I
Hi, 417 samples are required for 1 degree, and the frequency 621 is approximately 52,5 Hz, which is 1
Requires 353 samples and frequency 620 is approximately 32
, 7 Hz, requiring 1346 samples, and the frequency 622 is approximately 32, 8 Hz.
Therefore, 1359 samples are required. Similarly, using the frequency selection method shown in Figure 11,
If the frequency deviation is 0.3 or so, the frequency 711 is approximately 3
0,8 Hz, requiring 1424 samples, and frequency 712, approximately 31,0 Hz, requiring 1420 samples,
is about 32,6) (Z), which requires 1350 samples, and the frequency 722 is about 32,
8H2, requiring 1342 samples. The data rate benefit of the present invention can be shown by some simple calculations. Using the frequency selection method shown in FIG. 10, for a musical scale of 10 octaves, each is divided into three ρ tones. Therefore, 936 frequencies spanning 4.1 octaves were selected, for a total of 360 frequencies.
frequencies are selected. For a two-channel stereo system, there are 720 complex discrete Fourier transform coefficients of 10 pits each (6 bits for logarithmic magnitude and 4 bits for angle). These complex discrete Fourier transform coefficients have a principal octave of approximately 2
Assuming it starts at 0 Hz, about 20 sh per second
will be calculated and sent.

Due to this, the total data rate is approximately 14,400 sys per second.
It becomes 0 bit. This is an order of magnitude lower than the speed of 230,000 to 1,400,000 bits per second required by the prior art. This single-digit data rate reduction
It is believed that the techniques of the present invention, even taking into account the additional computations required by the present invention, will be very effective in addressing the requirements of a wide variety of memory devices and communication bands.

The present invention can be applied to both a 2-channel stereo system and a 4-channel stereo system without major changes.
It will be clear to those skilled in the art that they can be equally utilized. It will also be appreciated that some parts of encoder device 100 and decode device 200 can be used in common in a multi-channel device with appropriate time division data separation.

For example, the functionality of look-up tables 232 and 235 can be combined with appropriate address operations. Similarly, look-up tables 262 and 267
You can also summarize.

FIG. 12 shows a preferred embodiment for creating inverse filter coefficients for canceling room acoustic effects in a music signal reproduction device. FIG. 12 shows a memory 160 as shown in FIG. 1, which is connected to a decoding device 200 substantially as shown in FIG.

This decoding device 200 includes a digital filter 210 and an antilog. , a polar coordinate-orthogonal coordinate conversion means 230, an inverse Fourier transformation means 240,
A digital-to-analog conversion means 250 is provided. The digital-to-analog conversion means 250 is connected to a sound reproduction means such as the speaker 801 shown in FIG.

Speaker 801 is placed in a working position within room 802. Also within this room 802 is a microphone 803 placed in a preferred listening position. This microphone 80
3 is connected to an encoder P device 100 substantially as shown in FIG. The δ encoding device 100 includes an analog-digital conversion means 110, a Fourier transformation means 120, a rectangular coordinate-polar coordinate conversion means 130, and a logarithmic conversion means 140.

The output of the logarithmic conversion means 140 is connected to a processor 804 which operates as described below.

Memory 160 stores a specific test program that allows room 8020 room acoustics to be canceled by the apparatus shown in FIG. This type of test data typically includes one or more signals containing substantial acoustic energy across the frequency spectrum of interest.

In addition to recording test data in the memory 160, a test signal source 805 that generates such a test signal can also be connected to the digital filter 210.

Preferably, the test signal source 805 generates a series of identical digital data words. These digital data words are recognized by the decoding device 200 as complex discrete 7 Nilliers transform coefficients in the form of polar coordinates, so that when presented with these same “test words” from the test signal source 805, each of the decoding devices 200 This test signal is transmitted by speaker 801 to listening room 802.

Microphone 803 receives a test signal modified by the acoustics of listening room 802 . This signal is processed by the encoding device 100 and logarithmic conversion means 140
The output of is sent to processor 804. If the DEST signal tones have the same energy at each selected frequency, then if there is a deviation from the same energy at the output of the logarithmic conversion means 140, this is due to acoustic effects in the room. It is. Processor 804 receives the complex discrete Fourier transform coefficients in polar coordinate form from logarithmic lookup table 140 and calculates the reciprocal of its magnitude part for each selected frequency. The magnitude portion of this set of reciprocals is input to lookup table 214, which is part of digital filter 210. in this case,
Lookup table 214 must be either random access memory (RAM) or erasable programmable read-only memory (EPROM).

In this way, specific frequency coefficients are applied to the digital filter 2.
10, the decoding device 200 causes the speaker 801 to send a signal different from the one sent to the digital filter 210 to the listening room 802. In this case, when the contents of the lookup table 214 are specified as described above, the acoustic characteristics of the listening room 802 are canceled by the function of the digital filter 210. Thus, for example, if there is a 'dead spot' of a particular frequency at the location of the microphone 803 in the listening room 802, this will be sensed by the apparatus of FIG.
10 appropriate corrections are made.

This technique can equally be used in the case of a "hot spot 1" where the signal amplitude of a specific frequency is increased at the position of the microphone 803 due to the acoustic characteristics of the recording room 802.

FIG. 13 shows an alternative embodiment capable of performing the functions of the encoding device 100 shown in FIG. 1 and/or the decoding device 200 shown in FIG. The device 900 shown in FIG. 13 comprises analog-to-digital conversion means 110 as shown in FIG. The data words formed by this analog-to-digital conversion means 110 are sent to input controller means 901 and from there to an arithmetic logic unit 902. , this arithmetic and logic unit 902 performs logic functions under the control of a program permanently stored in read-only memory 903 . The read-only memory 903 is an arithmetic logic unit) 9'02 is a Fourier transform means 120. Cartesian coordinates −
It may be permanently programmed to perform the functions of the polar coordinate transformation means 130, the logarithmic transformation means 140, and the digital filter 150;
2 is a digital filter 21o1 and an antilog conversion means 22o.

Under the control of a program stored in the read-only memory 903, this arithmetic logic unit 902 may be programmed to perform the functions of the polar coordinate to rectangular coordinate transformation means 230 and the inverse Fourier transformation means 240. The arithmetic and logic unit 902 may periodically access intermediate processing data stored in the random access memory 904 based on the program stored in the read-only memory 903. This intermediate processing data is further calculated. The result of the logical operation of the arithmetic logic unit) 902 is sent to the output means 905 and from there to the digital-to-analog conversion means 280.

Since this device requires a large amount of calculation, the arithmetic
The unit 902 must be a very high speed device that performs the above processing of the music signal No. 1 data in real time. However, in such an embodiment, it is possible to quickly change the specific algorithm to be performed by simply changing the program stored in the only memory 903, and in particular, in the digital filter 150 and the digital filter 210. filter coefficients can be quickly changed. In addition, the arithmetic and logic unit 902 can be made responsive to a program created by a user and stored in a portion of the random access memory 904. Therefore, a user can change the filter coefficients in digital filter 150 or digital filter 210 by simply changing the values stored in random access memory 904. Accordingly, the fifth device 900 shown in FIG. 13 is a highly flexible digital signal processing device.

[Brief explanation of drawings]

FIG. 1 is a block diagram showing an encoding device according to the present invention; FIGS. 2, 3, and 4 are block diagrams showing another embodiment of the digitizer 114 shown in FIG. 1; and FIG. FIG. 6 is a block diagram illustrating a decode 9 device according to the invention; FIG. 6 is a block diagram illustrating a polar coordinate-orthogonal coordinate conversion and anti-log conversion means replacing the anti-log conversion means 220 and the polar coordinate-orthogonal coordinate conversion means 230 shown in FIG. , FIG. 7 is a signal diagram showing the same signal in both the time domain and the frequency domain, FIG. 8 is a frequency spectrum diagram including as an example selected frequencies from the 7-tier transform means, and FIG. 9 is a Fourier transform diagram. FIG. 10 is a frequency spectrum diagram according to another embodiment of the invention in which the selected frequencies of the means correspond to musical notes; FIG. Figure 11 shows a part of the same wavenumber spectrum in which the two selected frequencies of the Fourier transform means are contained within the envelope associated with each note. FIG. 12 shows a preferred embodiment of the apparatus according to the invention for producing inverse filter coefficients for canceling room acoustics, and FIG. 13 shows a largely programmable digital signal processing system according to the invention. 1 illustrates an embodiment of the invention as implemented by a digital computer; FIG. 100... Encoding device, 101... Analog input, 110... Analog-digital converter, 11
1 river low/? Sfilter, 112 Sample and hold device, 113... Clock, 114... TeJ type, 120... Fourier transform means, 121° 122.1
23... Discrete Fourier transformer, 130... Cartesian coordinate-polar coordinate conversion means, 131, 132... Multiplier, 133... Adder, 134.137... Address means, 135'; Square root lookup table, 136... Divider 1138... Arctangent lookup tapes, 140... Logarithm conversion means, 141... Address means, 142... Logarithm lookup table, 150... ...Digital filter, 151...Adder, 152...Counter, 153
. . . address means, 154 . . . frequency coefficient look-up table, 160 . . . memory, 170 . . . communication channel, 200 .
°Digital filter, 211... Adder, 212...
- Counter, 213...Address means, 214...
Frequency coefficient lookup table, 220... Antilog number conversion means, 221... Addressing means, 222... Antilog number lookup table, 230... Polar coordinate-orthogonal coordinate conversion means, 231.234...・Address means, 2
32...Cosine lookup table, 233.
236... Multiplier, 235... Sign lookup table, 240... Inverse Fourier transform means, 241, 242, 243... Discrete Fourier transformer, 250... Digital-analog conversion means, 251・
. . . Register, 252, 253. 254 . . . Current drive, 255 . 266. 264.269...Addresser, 262,267°2
65.270... Lookup tape, 263°268... Adder. Fig, 3 F7g, 4 Fig, 7 /'-, IM2

Claims (9)

[Claims] (1) analog-to-digital conversion means for receiving and converting an analog signal representing a music signal into a plurality of corresponding first digital data words; each representing the amplitude of said musical signal at a corresponding e/tar pal among a plurality of sample/gain tar pals, each of which is connected to said analog-to-digital converting means, and is connected to said analog-to-digital converting means for successive successive sets of selected sets. Fourier transform means for converting the first digital data word into a corresponding set of a plurality of second digital data words, each of said second digital data words having a dominant frequency and an octave relative thereto. represents a discrete Fourier transform means of complex numbers at corresponding frequencies in at least one set of frequencies, each containing only related frequencies, and is connected to the Fourier transform means described above and is transmitted from it. communication means for transmitting digital data words; inverse Fourier transform means connected to said communication means for transforming the set of digital data words sent by said communication means into successive sets of fifth digital data words; , each of said fifth digital data words being representative of the amplitude of the analog signal at a corresponding interval of the plurality of sampling intervals, and connected to said inverse Fourier transform means;
1. A music signal communication device comprising digital-to-analog conversion means for converting a plurality of third digital artwords into corresponding analog signals. (2) being connected between said Fourier transform means and said communication means, wherein the complex discrete Fourier transform coefficients are configured to transmit each of said second digital data words in polar coordinate form having a magnitude portion and an angular portion; Cartesian-to-polar coordinate transformation means are connected between the communication means and the inverse Fourier transformation means to convert each digital data word retrieved from the memory means into complex discrete Fourier transformation coefficients. Polar coordinates converted into Cartesian coordinate form with part and imaginary part −
Claim No. (1) comprising orthogonal coordinate conversion means
A music signal communication device according to paragraph 1. (3) The orthogonal coordinate to polar coordinate conversion means converts each second digital data word into a polar coordinate form having a magnitude portion having a first predetermined number of bits and an angular portion having a second predetermined number of bits. Claim (2) further comprising means for converting the first predetermined number, wherein the first predetermined number is larger than the second predetermined number.
A music signal communication device according to paragraph 1. (4) a logarithmic conversion means connected between the orthogonal coordinate to polar coordinate conversion means and the communication means, for converting the magnitude portion of each second digital data word into its logarithm; and antilog conversion means connected between the means and the polar to rectangular conversion means for converting the magnitude portion of each digital data word read from the memory into its antilog. A music signal communication device according to scope item (2). (5) being connected between said orthogonal-polar coordinate conversion means and said communication means, said means for converting each second digital data word P by a selected amount related to the corresponding frequency of said second digital data word; A music signal communication device according to claim (2), comprising filter means for correcting the magnitude portion. (6) connected between the communication means and the polar-to-cartesian conversion means, the size of each second digital data word being a selected amount related to the corresponding frequency of the second digital data word; A music signal communication device according to claim (2), comprising filter means for correcting the edge. 1-Hll (7) The Fourier transform means comprises a plurality of discrete Fourier transform means corresponding to each of a plurality of main frequencies,
The discrete Fourier transform means of each processor are responsive to successive sets of a predetermined number of first desotal data words 1, said predetermined number being equal to the reciprocal of the product of the dominant frequency and the length of the sampler interval; The conversion means is the first of the above set.
Claim (1) Transforming the digital data word into a second digital data word representing each of said dominant frequencies and complex discrete Fourier transform coefficients at said frequencies having an octave relationship thereto. The music signal communication device described above. (8) The music signal communication device according to claim (1), wherein the main frequency in each set of frequencies corresponds to a selected musical note. (9) The at least one set of frequencies is comprised of 12 sets of frequencies, and the main frequency in each set of frequencies corresponds to a different note within the main octave. Music signal communication device. 01 The main frequency in each set of frequencies is 8, and the main frequency in claim No. 8 is within the predetermined frequency range of the musical notes associated therewith.
1) The music-signal communication device according to item 1). (111) The music signal communication device according to claim α1, wherein each note of the main octave is combined with a predetermined number of main frequencies. A music signal communication device according to claim 19, wherein a first main frequency that is lower by a certain percentage and a second main frequency that is higher than the frequency of the note by the predetermined percentage are combined. Each note in the main octave has a first principal frequency that is a predetermined percentage below the frequency of that note, a second principal frequency that is equal to the frequency of that note, and a second principal frequency that is above the frequency of that note by the predetermined percentage. A music signal communication device according to claim (1), in which a third main frequency is combined with a third main frequency. A music signal communication device according to claim (1).Q51 Receiving an analog signal representing a music signal and converting the analog signal into a plurality of corresponding first digital data words. wherein each of said first digital data words corresponds to a corresponding interval/4 of the plurality of sampling intervals.
and a Fourier transformer for converting successive selected sets of successive first digital data words into corresponding sets of a plurality of second digital data words. converting means, each of said second digital data words having a main frequency and a frequency having an octave relationship thereto; encoder P(1G+) of a music signal, characterized in that it represents discrete Fourier transform coefficients of a complex number at corresponding frequencies among at least one set of frequencies, each of which includes: In polar coordinate form, the discrete Fourier transform coefficients have a magnitude part and an angular part,
9. A music signal encoding device as claimed in claim .alpha.9, comprising rectangular to polar coordinate conversion means for converting each second digital data word. (ID) The above-mentioned orthogonal coordinate-polar coordinate conversion means converts each second digital data word into a polar coordinate format having a size section having a first predetermined number of bits and an angular section having a second predetermined number of pits. and wherein the first predetermined number is larger than the second predetermined number.081 Between the orthogonal coordinate-polar coordinate conversion means and the communication means. 10. An apparatus for encoding a musical signal according to claim 10, comprising logarithmic conversion means connected to a logarithm of each second digital data word for converting the magnitude part of each second digital data word into its logarithm. □ Filter means connected to the coordinate-to-polar transformation means for modifying the magnitude portion of each second digital data word by a selected amount related to the corresponding frequency of the second digital data word. A music signal encoding device according to claim 1υ.
Each discrete Fourier transform means is responsive to a predetermined number of first digital data words of successive sets, said predetermined number being equal to the reciprocal of the product of the dominant frequency and the length of the sampling interval. Fourier transform means transform said set of first digital data words into second digital data representing complex discrete Fourier transform coefficients at said respective dominant frequencies and said 'frequencies having an octopus relation thereto. An encoding device for a music signal according to claim 1, which converts the music signal into a word. C9. The music signal encoding device according to claim α9, wherein the main frequency in each set of frequencies corresponds to a selected musical note. 7. The apparatus for encoding a music signal according to claim 1, wherein the at least one set of frequencies consists of 12 sets of frequencies, and the middle main frequency of each set of frequencies corresponds to a different note of the main octave. An apparatus for encoding a musical signal according to claim 1, wherein the dominant frequency in each set of frequencies is within a predetermined frequency range of the musical note associated therewith. @The music signal encoding device according to claim 1, wherein each note of the main octave is combined with a predetermined number of main frequencies. (to) Each note of the main octave is combined with a first main frequency force that is lower by a predetermined percentage than the frequency of the note. 1. Each note in the main octave is combined with a first main frequency that is a predetermined percentage lower than the frequency of that note, a second main frequency that is equal to the frequency of that note, and 3 main frequencies. A music signal encoding device according to claim 0. @ Claim 151, in which the main octave is combined with a predetermined number of main frequencies arranged at frequency intervals of the same ratio. A device for encoding a musical signal as described in the above.1) A second digital data word received from the Fourier transform means is connected to the Fourier transform means and connected to the memory means so as to store the second digital data word in the memory means. An encoding device for a music signal according to claim Qfj, comprising a memory writing means capable of recording a first digital data. and by processing each first digital data word P as complex discrete Fourier transform coefficients at corresponding frequencies in at least one set of frequencies, each containing only frequencies having an octave relationship thereto. comprising inverse Fourier transform means for converting a selected set of successive first digital data words P into a corresponding set of second digital data words, each said second digital data word P having a plurality of samplings.゛represents the amplitude of the analog signal at the corresponding interval f in the interval, and is connected to the inverse Fourier transform arm described above,
A device for decoding a digital data word 9 representing a music signal, characterized in that it comprises digital-to-analog conversion means for converting the second digital data word r of said set into a corresponding analog signal. ω connected between the receiving means and the inverse Fourier transform means, each receiver converting the received first digital data word from the polar coordinate form into complex discrete Fourier transform coefficients which transform the real and imaginary parts; A decoding device for a digital data word expressing a music signal as claimed in claim 1, comprising polar coordinate-to-orthogonal coordinate conversion means for converting the digital data into a rectangular coordinate form. 3υ The patent further comprises antilog converting means connected between the receiving means and the polar to rectangular coordinate means, each receiver converting the magnitude portion of the received first digital data word into its antilog. Apparatus for decoding digital data words representing a music signal as claimed in claim 1. C10 connected between the receiving means and the polar to rectangular coordinate converting means, each receiver receiving a first digital signal by a selected amount related to the corresponding frequency of the first digital data tower V; Device for decoding a digital data word representing a music signal as claimed in claim ω, comprising filter means for modifying the magnitude part of the digital data word P.關 Apparatus for decoding a digital data wave representing a musical signal as claimed in claim 1, wherein the main frequency in each set of frequencies corresponds to a selected musical note. The at least one set of frequencies comprises 12 sets of frequencies, and the main frequency in each set of frequencies corresponds to a separate note of the main octave. A decoding device for representing digital data words. Apparatus for decoding digital data words representing a musical signal as claimed in claim 1, wherein the dominant frequency in each set of frequencies is within a predetermined frequency range of its associated musical note. Apparatus for decoding digital data words representing a musical signal according to claim C351, wherein each note of the main octave is associated with a predetermined number of main frequencies. CI? ) Decoding of a digital data word representing a musical signal according to claim 1, wherein each note of the main octave is associated with a first main frequency that is a predetermined percentage lower than the frequency of that note. Device. (To) Each note in the main octave has a first principal frequency that is a predetermined percentage lower than the frequency of that note, a second principal frequency that is equal to the frequency of that note, and a frequency that is also lower than the frequency of that note. A device for decoding a digital data word representing a music signal according to claim 1, wherein the decoding device is combined with a third main frequency which is higher by a predetermined frequency. A device for decoding digital data words representing a music signal as claimed in claim 1, wherein the C31 main octave is combined with a predetermined number of main frequencies arranged at equal frequency intervals. <4 (I) A patent comprising memory reading means connected to the above-mentioned receiving means and connectable to the memory means so as to read the digital data word stored in the memory means and transmit it to the above-mentioned receiver, stage. A decoding device for a digital data word f representing a music signal according to claim 1.